An efficient and privacy-preserving ranked fuzzy keywords search over encrypted cloud data

As cloud computing becomes widespread, more and more users prefer to outsource their local sensitive data into the cloud. In order to protect data privacy, these sensitive data usually has to be encrypted before outsourcing, which makes effective data utilization a very difficult task. Although traditional searchable encryption techniques allow users to securely search over encrypted cloud data, they only support exact single keyword search, i.e. they do not allow any minor spelling errors or format inconsistencies. Besides, these traditional schemes support only Boolean search, without capturing any relevance of data files and rarely sort the search result. Recently, fuzzy keyword search over encrypted data techniques are introdeced to resolve the problem of spelling errors and format inconsistencis. However, they may incur large index size, search result inaccuracy and high search complexity, which greatly reduce the system usability and efficiency. In this paper, we propose the solution for privacy preserving ranked fuzzy keyword search over encrypted cloud data with small index. We use k-grams and Jaccard coefficient to constrcuct fuzzy keyword set and produce fuzzy results, and efficient relevance criteria (e.g., TF × IDF) to capture the relevance between data files and search requests. Extensive experiment result shows the efficiency of proposed scheme.